Received: March 19, 2017 426

DTNH Indexing Method:

Past Present and Future Data Prediction for Spatio-Temporal Data

John Ayeelyan1*, Sugumarn Muthukumarasamy2, Rengan Sivagurunathan Rajesh2

1 2*
Department of Computer Science and Engineering,
Manonmaniunm Sundaranar University, Thirunelveli, Tamilnadu, 627012, India.
2
Department of Computer Science and Engineering,
Pondicherry Engineering College, Puducherry, 605014, India
* Corresponding authors Email: johnmtech@gmail.com

Abstract: Indexing methods are developed to effectively process user queries in many real-time and moving object
management applications. The existing spatial data updating indexing methods are based on the Integrated binary Tree,
R-Tree, R*-Tree, Oct-Tree, Quad-Tree, Grid-Tree and Hex-Tree. The depth of these trees is unbalanced and
overlapping, hence the performance is reduced in the multi-structure indexing methods. D-Tree (Decompose - Tree)
based multi-structure spatio-temporal index method is proposed to find the present, past and future data. The new multi-
structural model called DTNH-Tree used to find the present, past and future data. It consists of D-Tree, TB*-Tree, NT-
Tree and hash table. The D-Tree indexing is used to get the spatial data and manage the moving objects in the road
network. A set of TB*-Tree is used to index the history of moving object on road networks. A set of NT Trees is used
to manage the current position of the recently updated data and find present data of the moving objects. NT-Tree indexes
the present and future information of the moving objects. Finally the set of hash tables is used for updating the data
continuously. The proposed multi-structure indexing method supports different types of query processing compared to
the existing indexing methods. Experimental results exhibits better updation and query performance compared to the
MSMON-Tree and PPF*-Tree.
Keywords: Spatio-temporal data, Indexing methods, Query processing.

handle and predict moving objects database system.

1. Introduction The past, present and future data Predictions in
moving objects data is complex one in spatial (space)
Nowadays query processing and indexing
- temporal (time) environments because it changes
methods have an essential one for data retrieval in
the space and time in each and every seconds. The
spatial - temporal network. Query processing is a
numerous research effect taken to solve the issues of
collection of data arranged for ease and speed of
continuous moving objects databases in indexing and
search and retrieval of data. Indexing is a key data
query processing. The spatio-temporal indexing
structure used to retrieve data from database tables
methods are categorized into three types [1, 2] first
and improve the speed of data retrieval in movement
method deal about spatial index access methods , the
database. The indexing methods and query
second method is spatio-temporal index access
processing is inter-related one, based on the indexing
methods , the third is indexing the temporal
method different types of query processing types are
dimensions, while in this method spatial data is
supported to the particular indexing method.
considered as the second priority. The third method
Nowadays developing efficient indexing and query
is used to predict all the point of the data and the data
processing methods is an important research in
prediction based on the temporal and based on spatial
Moving object database. The traditional moving
is very complex because this type of indexing are
objects database (MOD) is not well equipped to
multi-structural and multi-model.it consists of more
International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017 DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017
than two indexing methods is used to processing all
the point of the data. The some of the main spatial
indexing are RT-TREE [3], 3D-R-Tree [4], STR-
Tree [5], MTSB-Tree [6], FNR [7], MON [8]. The
spatio-temporal index methods MR-Tree [3], HR-
Tree[9], HR+-Tree[10], MV3R-Tree[11]. The
Temporal dimension with spatial priority indexing
methods are TB-Tree[12], SETI [13], SEB-Tree [14],
PMR-Quad tree[15], TPR-Tree [16], TPR*-Tree [17],
LU Grid[18], RUM-Tree [19], IMORS [20] and
present research direction indexing methods Rppf-
Tree [21], PCFI+-Index [22], BBx-Index [23], UTR-
Tree [24], STCB+-Tree [25], PPFI-Index [26].The
temporal based spatial indexing is used index all the
points of indexing methods that is index the past,
present and future data prediction of indexing method.
The moving objects can be grouped into three
movement scenarios [27], namely constrained,
unconstrained, and movement in transportation
networks. The movement scenarios are considered as
different space than compared to first two scenarios.
This paper is based on movement in transportation
networks and third method in spatio-temporal
indexing method.
The Most of the spatial data updating indexing
methods [28] are based on the integrated binary tree,
R-tree, R*-Tree, Oct tree, Quad tree, Grid tree and
Hex tree. Those derived from R-tree employ a
minimum bounding rectangle (MBR) or a time-
parameterized minimum-bounding rectangle
(TPMBR) to represent a moving object. The range of
the bounding rectangle is always larger than the range
of the moving object itself. This is the fundamental
reason why they cannot resolve the overlapping
problem in high dimensions. The depth of these trees
is unbalanced and overlapping, hence the
performance is reduced in the multi-structure
indexing methods. To Reduce these problem new
indexing method is D-Tree (Decompose-Tree) based
multi-structure spatio-temporal index method is
proposed to find the present, past and future data. The
proposed multi-structure indexing method supports
different types of query processing compared to the
existing indexing methods. Experimental results
exhibits better updation and query performance
compared to the RPPF-Tree, STCB+-Tree, MSMON-
Tree and PPF* -Tree.
The organization of this paper is as follows.
Section 2 is related works and motivation and section
3 describes data model and data structure of the new
hybrid index method. Section 4 reports the
performance evaluation, and conclusion is given in
section 5.
International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017
427
2. Related work
In this section, gives the brief literature about
various saptio-temporal access methods for past,
present and future indexing methods. These indexing
methods come under the categories of spatio-
temporal access method or indexing at all the point of
time. Past-Current-Future+-Index (PCFI+-Index)
[22] indexes the past, current and future information
of the dynamic objects and the finds an object's
position at every point of time. The PCFI+-Index
constructed based on the PCFI-Index, which was
based on SETI-tree and TPR*-tree. The PCFI+-Index
having memory parts with the name frontline, and
disk based part. The spatial access is indexed using
region partitioned into the non-overlapping cell.
TPR*-tree index the moving belong the on
overlapping cells. The hash table is indexed large
updating and find the current position of the moving
objects. The PCFI+-Index support most of the queries
efficiently and provide only uniform solutions for the
trajectory query. BBx Index Method [23] used to find
queries about the past, present and future position of
moving objects. BBx index structure is used to find
the position of moving objects in linear function of
time. BBx index inherits the ability of finding present
and future position of the BBx-Tree. Three factors are
considered to the stage efficiency are Object position
is represented as a linear function. It is done by TPR-
tree family (spatial representation), Object is based
on indexing BBX, B+-Tree used for updation. RPPF-
Tree is used to [21] obtain all the online position
information from the moving e-service users. The
proposed work offers an RPPF-Tree technique capable
of capturing moving objects at all positions in time.
This tree supports only valid transaction time to
monitor the applications and it supports time related
query and position related query. VPMR-Tree [27]
indexing is used to provide the solution to
approximate future information for moving objects
and indexing the road network moving objects. This
indexing method is a variation of PMR Quad tree and
real time data prediction purpose dynamic array is
used. Past, Present and Future Index method (PPFI)
[26] is used to predict the current and near future
position of moving objects on a fixed road network
with an efficient update mechanism. The proposed
work is PPFI and which consists of a 2DR*-Tree,
1DR*-Tree and hash table for updating. 2DR*-Tree
is a static part built on managing the fixed network.
1DR*-Tree is a dynamic part built on indexing object
movements. This indexing method is used R Tree and
R* Tree so data overlapping problems occur. Final
Future Trajectory Prediction Algorithm (FFTPA)
[29] method is based on the R-Tree and its support
DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017 428

related to past position and trajectories. It has a 3. Proposed Method

dynamic road network and efficient update
mechanism. 3.1 Problem analysis and data model
The FT-Tree Index method is used to predict past,
present and future prediction of the moving objects The past, present and future indexing hybrid
[30]. The moving object index for full temporal query, methods are based on the integrated binary tree, R-
proposes a moving object index structure FT Tree tree, R*-Tree, Oct tree, Quad tree, Grid tree and Hex
supports for temporal query, it is based on TPR-Tree. tree. The main problem is having dead spaces and
In this FT-Tree is used TPR so it is based on the R- overlapping problems are occurring in the moving
Tree and R Tree based indexing data have objects updation. The data overlapping produce
overlapping problems are occurring. HTPR*-Tree massive data updation, so the data retrieval time is
Indexing [31] is the foundation for the past, present more and it produces worse performance. The
and future and used to predictive queries, but also MSMON- Tree [32] addressed this problem, in this
partial history ones involved from the most recent model also used R* -Tree and TPR-Tree is stored
update. This index structure, named History Time- spatial data information. The structure of the R-Tree
Parameterized R-tree (HTPR*-tree), which takes into and R*- Tree shown in the Figs. 1 and 2.
account moving object creation time or update time
in the leaf node entry, and supports partial history
query. This indexing method also based on the R-
Tree and R*-Tree. MSMON -Tree [32] is used to
retrieve the past, present and future position of
network constrained moving objects. FNR & MON
tree only deal with history of data and cannot support
the query of moving objects in the current and future
positions. The structure of the hybrid tree is support
multi-Grid-R*-Tree used to index the road network.
R*-Tree is used in a route in the road network and
also find trajectories of the moving objects in the road.
PPFN*-Tree [33] Indexing of moving objects
grouped into two categories are deal with past Figure.1 R-Tree
information retrieval and future data prediction. It is
hybrid structure which consists of 2DR* managing
road networks and set of TB*-tree indexing the object
movement history trajectories along the polyline.
HTPR* indexing the position of the moving object
after recent updates. Update improvement purpose
hash table is used with 2DR*-Tree. The previous
indexing methods of spatial data updating indexing
methods are based on the integrated binary Tree, R-
Tree, R*-Tree, Oct-Tree, Quad-Tree, Grid-Tree and
Hex-Tree. The depth of these trees is unbalanced and Figure.2 R*-Tree
overlapping, hence the performance is reduced in the
multi-structure indexing methods. The various types
of indexing and the query process sing are mentioned
in [38] and further consideration in spatio-temporal is
security mentioned in [34-37] these papers are
mentioned various advanced security features. The
PPFI [26], MSMON [32], PPFN*[33] and D-Tree
[28] are the main past, present and future data
prediction indexing methods and these methods are
considered as comparison purpose.

Figure.3 Road Segment Representation

International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017 DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017
Figure.4 Proposed index method
In the moving object data representation, network
model is necessarily one. The proposed network
model represented as in term of routes, junctions and
roads (i.e) network G = (P, R), where P = {P1..Pn}
is a set of junction, R is the route and A, B, C, D is
the points and S1, S2, S3 is the segment between two
points. The A and B is the starting and ending point
between the junctions. The sequence of road segment
in the continuous network has shown in the Figure 3.
3.2 Structure
The proposed multi- model structure is called
DTNH Tree shown in the figure 4. This is a multi
model structure consists of D-Tree (D-Compose
Tree), TB*-Tree (Trajectories Bundle Tree), NT-
Tree (Networks -Limited TPR-Tree), and hash table.
D-Tree is used to answer the spatial information such
as geometric information (i.e) Road information, data
transformation, actual time and spatial location of
moving objects noted directly. In D-Tree spatial
information noted directly so time complexity is
reduced, so performance automatically increased.
TB*-Tree used to find trajectories of moving objects
and the NT-Tree is used to find the present and future
position of moving objects. Hash Tables are used for
updation purpose. The Hash tables are connected
each tree node so it is easily updates the data.
The proposed DTNH-Tree indexing has three
parts. The first part is Decompose Tree (D-Tree). The
D-Tree is a space partitioning method for accessing
moving object data. In this part all the spatial points
International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017
429
are noted with time attributes. The main attributes are
[ID, T, S, A], where ID is a unique object identifier, T
is a continuous time stamp, S is a spatial Co-ordinates,
A is a time attributes. Using this D-Tree recursively
decompose spatial data with time attributes. So easily
get all the records in the, spatial coordinate wise and
time wise moving object data is update in the table.
This D-Tree is based on partitioning, so data
overlapping is avoided, easily access the data and
time complexity is reduced. Finally The D-Tree is
linked to a hash table for continuous updations. The
Second part is TB*-tree indexing having moving
object history trajectory along with position of
moving objects. This TB*-Tree is linked with hash
table and D-Tree. TB*-Tree manage both spatial and
Temporal informations. The third part is NT-Tree is
designed and manages moving object information.
Queries about future position information are
retrieved from NT-Tree. This tree also updated with
hash table for current updation.
3.3 Insertion
Insertions of the moving objects are happening
with the help of Hash Tables (HT). P, R, S and O
information are located and then inserted to hash
table structure. The spatial information and temporal
data [ID, T, S, A] are inserted into the leaf node of the
hash table structure. The coming objected are inserted
into TB*-Tree with the help of hash table structure.
The present moving objects are moving from a hash
table to NT-Tree. All this hash table and NT-Tree and
TB*-Tree are implemented with the help of the
Pointer. The after insertion of hash table from the NT-
Tree and TB*-Tree data are moved to D-Compose
Tree. With the help of D-Tree data overlapping are
avoided.
ALGORITHM 1. Insert [ID, T, S, A]
/* input: P, R, S and O information inserted into
the hash table and S1 and S1 starting and ending point
of Road segments */
1. Get Route (R) information and moving
Object (O) are inserted
2. Invoke hash table
3. If (hash table, Object > 0) objects are
inserted
4. Else Hash table is empty
5. Check time duration then
6. If any objects are find then insert
7. Otherwise
END Insert.
DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017
3.4 Updation
The moving objects, updating are updated with
the help of hash table, D-Tree and TB*. The hash
table is used for dynamic updation and objects are
moving next indexing and the hash table is resize the
objects when updations are change. D-Tree is used to
update the road information. TB*-Tree update the
Road trajectories and moving update moving objects
updations in hash tables. First moving objects are
updated in a hash table and then spatial data are
moved to D-Tree and TB*-Tree. The present data
updated in NT Tree and here all the updations are
updated directly from hash table to D-Tree and TB*
and NT-Tree so performance is increased because
time complexity is reduced. The main updation
parameters are Moving Objects (MO), Moving
Object Identification (MO.id) Object Velocity (V),
Time (T), Segment (S), Position (P).Three Situation
is considered for updation of temporal data. First
Velocity (V)is considered for updation. Second Time
threshold (T) is considered and Third is moving
objects are moving in opposite direction is also
considered for updation. The main spatial updation is
new place or position find, new entry (N) is
considered and this new entry updated in D-Tree. The
updation algorithm is as follows.
Algorithm 2:Updation
Input: /*moving object information (MO, MO.id),
Velocity (V), Time (T) and spatial information (P, S,
N)*/
1. If hash table (Mo, MO.id)!=0;
2. MO position (P) is moved to D-Tree
3. Update hash table node;
4. If Objects opposite directions
5. Delete old node trajectory and new Entry (N)
intoTB*-Tree
6. Hash table node is moved into NT-Tree
7. Else delete from hash table
8. Search and insert new into TB*-Tree and
NT-Tree
9. Hash table (Mo, MO.id) point to store
moving object information
End Update.
3.5 Search procedure
DTNH -Tree indexing support to different types
of search procedure compared to other indexing
method because D-Tree is support to most of the all
spatial queries and remain two trees are supported for
temporal queries. And it also supports to spatio-
temporal queries also. The main Query types are
point query, Time slice query, spatial slice query,
window query, moving query, trajectory query,
International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017
430
Topological query and predictive query. TB* is
supports both spatial- temporal queries and NT-Tree
supports temporal queries.
Point Query
Point query is used to find exact information, in
particular spatial portion P at a time interval T. This
query retrieves all the nearest spatial information.
The representation of point query is Q = (P, T).
Time Slice Query
The time slice query retrieves spatial data in term
of time slice T and spatial rectangle R at the time
instance. B-Tree, R-Tree and R*-Tree are also used
to find spatial information, but the main drawback is
overlapping of rectangle and data. The proposed
work has rectified this problem because D-Compose
tree rectified this problem.
Spatial Slice Query
Spatial Position (P) informations are retrieved
particular time slice interval [Ts, Te].The nearest
information is retrieved in particular time. The
representation of spatial Slice query Q = (P, Ts, Te).
Window Query
Multi-dimensional spatial rectangle [R] area data
is valid during a particular time interval [Ts, Te] with
the help of D-Tree is overlapping is avoided. Query
representation is Q = (X1, Y1, X2, Y2, Ts, Te).
Temporal Query
This query is used to find time related
information which is retrieved with time interval. For
example, what is the object position of 10.30 am?
This query information is retrieved with the help of
velocity and time. The Query representation is Q =
V*T.
Spatio-Temporal Range Query
This query is used to find all the objects have
with certain Range R = (X1, Y1, X2, Y2) during time
interval T = (Ts, Te).
Topological Query
This query is used to find all the objects are in
enter, leave, cross a given position R = (X1, Y1, X2, Y2)
with time interval T = (Ts, Te).
DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017 431

Continuous updation Query

Continuous updation query is used to update all

the information. The updation is based on velocity
(V), Time (T) and position (P). The updation query is
all the events are updated with instance of time. The
query representation is Q = (V, T, and P)

Predictive Query

DTNH indexing method supports to predictive

query. The predictive query predicts the data based Figure.5 Update Cost Comparison
on recent update time Interval (T1), update position
(P1) and velocity (V). Based on this currently what
information available and next time interval
(T2), it predicts further position (P2).

4. Experimental setting and details

The evaluation performance of the DTNH -Tree
is compared with MS-MON-Tree [32] and PPFN*-
Tree [33]. The Experimental purpose randomly
generated, distributed network based moving object
are used. On real implementation is based on real
road network data is used as input. The position of the Figure.6 Query Efficiency Comparison
object and the velocity of the objects are considered
for experimental and data prediction purpose. The
position in X-Y coordination and moving objects are
considered in the form of (Object id, Position (P)),
where an object id denotes where the objects are
available and relative position of the objects are
denoted. After identifying the object position and
object attributes, the number of segments also noted.

4.1 Performance analysis

Update Cost Comparison

The figure 5 shown the compare with the update

cost of DTNHTree with MSMON-Tree and PPFN* Figure.7 Query Cost Comparison
-Tree. The CPU time is Increase when the increase
the number of updates. The main advantages of the
updation are node access performance is increased
because the node overlapping is avoided with the help
of D-Tree. The comparison purpose data are
generated from 0 to 1000 and updation is checked at
different times and different query ranges.

Figure.8 Trajectory Query Cost Comparison

International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017 DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017
Query Efficiency
Data are generated 0 to 1000 and different query
ranges are taken for checking query efficiency and
cost comparison. The query efficiency and cost
comparisons are compared with previous indexing
methods Shown in the Figure 6 and Figure 7. The
query performance is increased because R-Tree is
avoided and overlapping also avoided. The query
range is divided into segment wise so performance is
increased because within time search particular
segments. The Query range segment are often very
large the search time is also increased in MSMON-
Tree [32] and PPFN*-Tree [33]. Here it is avoided
because no overlapping. Trajectory data are updated
with help of TB*- Tree and the query performance is
increased because each segment spatial data are
updated with the help of the D-Tee and the trajectory
query performance shown in the Figure 8.
5. Conclusion
DTNH structure was proposed to index the
moving objects prediction in the past, present and
future positions The existing spatial data updating
indexing methods are based on the integrated binary
Tree, R-Tree, R*-Tree, Oct-Tree, Quad-Tree, Grid-
Tree and Hex-Tree. The depth of these trees is
unbalanced and overlapping, hence the performance
is reduced in the multi-structure indexing methods.
DTNH multi-structure spatio-temporal index method
is proposed to find the present, past and future data.
DTNH index method manages past, present and
future data and it predicts the data in different
position and time. This proposed multi-structure
indexing method supports different types of query
processing compared to the existing indexing
methods. Experimental results exhibits better
updation and query performance compared to the
MSMON-Tree and PPF*-Tree. DTNH index method
supports better query range in compared to MSMON
and PPFN*. The DTHN Tree base Tree Having D -
Tree it is used to avoid overlapping of data, so
performance of the index method is increased and
time complexity also decreased. The performance of
the query is compared to the previous indexing
methods it shown better performance in number of
queries supporting over previous indexing methods.
The further research includes different direction of
moving objects updation and predictions of data in
different time interval are considered.
International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017
432
References
[1] M. Mokbel, T. Ghanem, and W. G. Aref,
Spatio-temporal access methods, IEEE Data
Eng. Bull., Vol.26, No.2, pp. 4049, 2003.
[2] L.V. Nguyen-Dinh, W. G. Aref, and M. F.
Mokbel, Spatio-temporal access methods: Part
2 (2003 - 2010), IEEE Data Eng. Bull., Vol.33,
No.2, pp. 46-55, 2010.
[3] X. Xu, J. Han, and W. Lu, RT-tree: an
improved r-tree indexing structure for temporal
spatial databases, In: Proc. of the International
Symposium on Spatial Data Handling, SDH,
pp.10401049, 1990.
[4] Y. Theodoridis, M. Vazirgiannis, and T. Sellis,
Spatio-temporal indexing for large multimedia
applications, In: Proc. of the IEEE Conference
on Multimedia Computing and Systems, ICMCS,
1996.
[5] D. Pfoser, C. S. Jensen, and Y. Theodoridis,
Novel approaches in query processing for
moving object trajectories, In: Proc. of the
International Conf. on Very Large Data Bases,
VLDB, pp.395406, 2000.
[6] P. Zhou, D. Zhang, B. Salzberg, G. Cooperman,
and G. Kollios, Close pair queries in moving
object databases, In: Proc. of the GIS, pp.211,
2005.
[7] E. Frentzos, Indexing objects moving on fixed
networks, In: Proc. of the SSTD, pp.289305,
2003.
[8] V. T. De Almeida and R. H. Gunting, Indexing
the trajectories of moving objects in networks,
Geoinformatica, Vol. 9, No.1, pp.3360, 2005.
[9] M. Nascimento and J. Suva, "Towards historical
rtrees", In: Proc. of the ACM Symposium on
Applied Computing (ACM-SAC), pp. 235-240,
1998.
[10] Y. Tao and D. Papadias, "Efficient historical r-
trees," In: Proc. of the 13th IEEE Conference on
Scientific and Statistical Database Management
(SSDBM), pp. 223 -232, 2001.
[11] Y. Tao and D. Papadias, "MV3R-tree: a spatio-
temporal access method for timestamp and
interval queries", In: Proc. of the 27th
International Conference on Very Large Data
Bases, pp.431-440, 2001.
[12] D. Pfoser, C. S. Jensen, and Y. Theodoridis,
"Novel approaches in query processing for
moving object trajectories", In: Proc. of the 26th
International Conf on Very Large Data Bases,
pp.395-406, 2000.
[13] V. P. Chakka, A. Everspaugh, and J. M. Patel,
"Indexing large trajectory data sets with SETI",
DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017
In: Proc. of the Conf on Innovative Data Systems
Research, CIDR, 2003.
[14] Z. Song and N. Roussopoulos, "SEB-tree: an
approach to index continuously moving objects",
Mobile Data Management, MDM, pp. 340344,
2003.
[15] J. Tayeb, O. Ulusoy, and O. Wolfson, "A quad
tree-based dynamic attribute indexing method",
The Computer Journal, Vol. 41, No. 3, pp. 185
200, 1998.
[16] S. Saltenis, C. S. Jensen, S. T. Leutenegger, and
M. A. Lopez, "Indexing the positions of
continuously moving objects", In: Proc. of the
ACM International Conf. on Management of
Data, SIGMOD, pp. 331342, 2000.
[17] Y. Tao, D. Papadias, and J. Sun, "The TPR*-
tree: an optimized spatio-temporal access
method for predictive queries", In: Proc. of the
International Conf. On Very Large Data Bases,
VLDB, 2003.
[18] X. Xiong, M. Mokbel, and W. G. Aref,
"LUGrid: update-tolerant grid-based indexing
for moving objects", In: Proc. of the MDM, pp.
13, 2006.
[19] Y. N. Silva, X. Xiong, and W. G. Aref, "The
RUM-tree: supporting frequent updates in r-
trees using memos", VLDB J., Vol. 18, No. 3, pp.
719738, 2009.
[20] K.S. Kim, S.W. Kim, T.W. Kim, and K.-J. Li,
"Fast indexing and updating method for moving
objects on road networks", Web Information
Systems Engineering Workshops, pp. 3442,
2003.
[21] M. Pelanis, S. altenis, C.S. Jensen, "Indexing
the past, present, and anticipated future positions
of moving objects", ACM Transactions on
Database Systems (TODS), Vol. 31, No. 1, pp.
255-298, 2006.
[22] Z.H. Liu, X.L. Liu, J.W. Ge, and H.Y. Bae,
Indexing large moving objects from past to
future with PCFI+-index", In: Proc. of the
COMAD, pp. 131137, 2005.
[23] D. Lin, C. Jensen, B. Ooi, and S. Saltenis,
"Efficient indexing of the historical, present, and
future positions of moving objects", In: Proc. of
the MDM, pp. 5966, 2005.
[24] Z. Ding, "UTR-tree: an index structure for the
full uncertain trajectories of network-
constrained moving objects", In: Proc. of the
MDM, pp. 3340, 2008.
[25] H.Lin, "Indexing the trajectories of moving
objects", In: Proc. of the International Multi
Conf. of Engrs. and Computer Scientists, pp.
732737, 2009.
International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017
433
[26] Y. Fang, J. Cao, Y. Peng, and L. Wang,
"Indexing the past, present and future positions
of moving objects on fixed networks, In: Proc.
of the International Conf on Computer Science
and Software Engineering, pp. 524-527, 2008.
[27] D. Pfoser, Indexing the trajectories of moving
objects, IEEE Data Engineering Bulletin Vol.
25, No. 2, pp. 29, 2002.
[28] Z. He, C. Wu, G. Liu, Z. Zheng, and Y. Tian,
Decomposition tree: a spatio-temporal
indexing method for movement big data,
Journal Cluster Computing, Vol. 18, No. 4,
pp.1481 -1492, 2015.
[29] I. Sasikala, M. Ganesan, and A. John, Uncertain
data prediction on dynamic road network, In:
Proc. International Conf on Information
Communication and Embedded Systems
(ICICES), 2014.
[30] Y. Zheng, A fast index method for moving
objects on full temporal query, In: Proc. of the
3rd International Conf on computer Research
and Development (ICCRD), 2011.
[31] Y. Fang, J. Cao, J. Wang, Y. Peng, and W. Song,
"HTPR*-tree: an efficient index for moving
objects to support predictive query and partial
history query", Web-Age Information
Management Lecture Notes in Computer
Science, Vol. 7142, pp. 26-39, 2012.
[32] Z. Zhu, Q. Yang, and D. Pi, "Past, current and
future positions index of moving objects in
networks future", Control and Automation
Lecture Notes in Electrical Engineering, Vol.
172, pp. 335-342, 2012.
[33] Y. Fang, J. Cao, Y. Peng, N. Chen, and L.
Liu, "Efficient indexing of the past, present and
future positions of moving objects on road
network", Web-Age Information Management
Lecture Notes in Computer Science, Vol. 7901,
pp. 223-235, 2013.
[34] T.C. Segar and R. Vijayaragavan, Pell's RSA
key generation and its security analysis, In:
Proc. of the Computing, Communications and
Networking Technologies (ICCCNT), pp. 1-5,
2013.
[35] C. Thirumalai, Physicians drug encoding
system using an efficient and secured linear
public key cryptosystem (ESLPKC),
International journal of pharmacy and
technology, Vol. 8, No. 3, pp. 16296-16303,
2016.
[36] C. Thirumalai, and S. Shanmugam, Multi-key
distribution scheme using Diophantine form for
secure IoT communications, In: Proc. of the
IEEE IPACT, 2017.
DOI: 10.22266/ijies2017.0630.48
Received: March 19, 2017 434

[37] C. Thirumalai, and P. Viswanathan,

Diophantine based asymmetric cryptomata for
cloud confidentiality and blind signature
applications, JISA, Elsevier, 2017.
[38] A. John, M. Sugumaran, and R.S. Rajesh
Indexing and query processing techniques in
spatio-temporal data, Ictact Journal on Soft
Computing, Vol. 6, pp.1198-1217, 2016.

International Journal of Intelligent Engineering and Systems, Vol.10, No.3, 2017 DOI: 10.22266/ijies2017.0630.48